I think it's also important to give a new DM or an unfamiliar DM a fair shake--if you don't like a couple of decisions, stick around a little unless the type of game (and the people in it) aren't to your liking. Give constructive feedback. Say what you like and don't like and give it some time.
The players, of course, can walk away from draconian or unfair GMs. Start their own game. Find a new GM. They are not forced to play in the GM's game. However, the GM is indeed entitled to the power to say Yes, No, or Somewhat. That is really his only function. There are other things to consider, obviously, but rule-wise, that's his job. And, as far as I can tell, it's literally written in the book that the GM does indeed have this level of power over the game.
To the extent that there are differences here, I think I tend to lean RedTonic's way.
The frequent suggestion on this thread (not just from JamesonCourage but also Ariosto, I think RC, and maybe others I'm forgetting) that a player has the right to walk from a GM's game is true but not really helpful. We're here talking about what makes for better or worse GMing, across a range of editions of D&D (and some other games) and a range of playstyles. Given that purpose, I think there is a lot more to be said about
how to GM well than simply "Exercise whatever power you want to subject to the knowledge that if your players really hate it, they'll quit your game". After all, no creative writing course gives instructions saying "Write whatever you want, subject to the knowledge that you'll only make money from it if you find a publisher who thinks the public and/or critics will like it." Part of the point of a creative writing course is to
learn how to write stuff that will withstand critical scrutiny. Similarly, part of the point of this discussion is presumably to
share a range of ideas about how a GM can make the play experience a better one (with "better" being relative to a range of rulesets, playstyles etc).
For some rulesets, it's
just not true that the GM has the level of power over the game of being entitled to say "yes" or "no" to a player's call in respect of any of his/her PC's action. Now for some potential players, that might be a reason to avoid those rulesets, or to add such a rule into the way they play the game. But it's also interesting to think about
why a ruleset might be written which doesn't include such a rule, what sort of play experience it might be intended to promote, how that play experience might interact with other preferences and concerns at the table, etc.
For example, if a 4e GM purports to veto the Knocked-down Snake, and a player queries that, I don't think it's very helpful to just start talking about "pushy" or "whiny" players who don't understand the GM's role. What sort of play experience is the GM trying to promote. If it's about the coherence of the GM's fiction, what is the GM's understanding of the players' contribution to that fiction? Does the GM object to knocking a snake down per se, or to the fact that the PC can do it willy nilly when in the real world it would be very hard, or what? If the GM won't allow a player takeback, is that because of the effect of takebacks on immersion, or because the GM thinks that players should be bound by some sort of equivalent to chess's "touch, move" rule, or what?
Whether or not it would be productive to raise this sort of issue actually at the gaming table in question is one thing. But I can't see how it is remotely out of order to raise them on a discussion board. And once these sorts of questions are put into play, I can't see that the range of satisfactory, considered responses is exhausted by "Of course a GM has that sort of power, but equally players are free to quit the game". The actual history of RPG design and play shows that other things are possible!
EDITED TO ADD:
One of the premises of a creative writing course is that even if you know what you like when you see it, it can be non-trivial to
proudce that stuff yourself. Presumably the same is true for GMing. So while I agree that people should play what they like, I nevertheless think that critical reflection can help work out what it is that they're doing that is contributing to them getting what they like, and what it is that they're doing that is impeding them getting what they like. I will freely admit that my game has improved by becoming more self-conscious about my conception of the function of the rules, the GM, the way these can relate, the range of purposes they can serve, etc. Am I really an outlier? Is everyone else already running a game that is (by their own lights) the very best that it can be?